Spieth HR, Xue F, Strau K. (2004)
Induction and inhibition of diapause by the same photoperiod: experimental evidence for a "double circadian oscillator clock".
J Biol Rhythms. 2004 Dec;19(6):483-92.

On the southern Iberian Peninsula, the seasonal life history of the large white butterfly, Pieris brassicae, comprises 2 different photoperiodically induced developmental arrests: a hibernation diapause at photophases < 11 h and an estivation diapause at photophases > 14 h. At intermediate photophases (12 h to 13 h), the butterfly responds with a nondiapause. Combined with the experimental setup to determine photosensitivity in insects, the different photoperiodic responses at long-, intermediate-, and short-night conditions were examined to gain more insight into the time measurement mechanism in P. brassicae. The study reveals evidence for a "double circadian oscillator clock" mechanism that is based on 2 submechanisms, a "short-night determining system" and a separate "long-night determining system." This conclusion was drawn from the facts that an LD 9:15 long-night induces a hibernation diapause but inhibits an estivation diapause and, conversely, that an LD 16:8 short-night inhibits a hibernation diapause but induces an estivation diapause. This opposite effect of the same photoperiod supports the argument for the existence of 2 independent targets for light-dark cycles, interpreted as 2 antagonistic time measurement systems. The existence and independence of 2 systems was further shown by differences in long-night versus short-night responses regarding photosensitivity, temperature dependence, and heritable factors. The long-night measurement system is most effective in the 5th larval stage, is highly affected by temperature, and is easy to manipulate by selective inbreeding. The short-night measurement system is most effective in the 4th larval stage, is largely temperature compensated, and is not affected by experimental manipulation of the longnight measurement system.